Processing notes from README provided by ODF:
The CTD depths in the shipboard version (20 Dec 2013) of the data - bottle data in particular - were originally calculated incorrectly due to two software errors. A small error in the gravity calculation affected both bottle and CTD data. In addition, the bottle file CTDDEPTH calculation was inadvertently using a combination of latitude and longitude, vs just latitude, as input to the gravity component of the depth equation, causing up to a 20m error in some of the casts.

Two different depth calculations were subsequently added into these data files: the original, integrated Saunders-Mantyla integrated depth calculations (SMDEPTH), with the gravity error corrected; and the non-integrated Fofonoff-Millard depth calculations (FMDEPTH), same as those used by SBE. The Fofonoff-Millard depths are used for the primary CTDDEPTH column in both bottle data files.

ODF reports that the CTDDEPTH column agreed best with depths identified from CTD bottle trips recorded during GEOTRACES CTD acquisition. The depth values in CTDDEPTH column are from the Saunders & Fofonoff 1976 calculation of depth from pressure. This is the same default calculation of depth used by SeaBird (SBE) software. GEOTRACES CTD data acquisition uses SBE software. It follows that it would be preferable and consistent for GEOTRACES and ODF to use the values in the CTDDEPTH column.

Processing notes from README_1410.txt:
Separate columns of S-M and F-M depths ("SMDEPTH" and "FMDEPTH") are included in bottle and CTD files. In addition, "ODF_CTDPRS" is an added column in the bottle files. This may differ from the main CTDPRS in the bottle file only for EPZT GT-C casts, where Seasave-processed CTD trip data were used.

Note: 'FLAG_W' columns = WHP (WOCE Hydrographic Program) quality flags. ODF used the WHP quality coding schema, but also included two non-standard quality codes in the GT-C bottle file:
- QUALITY CODE 'A' WAS USED FOR DATA VALUES SUBMITTED AS 'ABOVE DETECTION LIMITS'
- QUALITY CODE 'B' WAS USED FOR DATA VALUES SUBMITTED AS 'BELOW DETECTION LIMITS'
This was at the request of two different PIs.